At Smart Sensing Ltd., we’re committed to revolutionizing the way cities and businesses gather and analyze data. Our proprietary AIoT architecture provides cutting-edge smart city and business intelligence solutions that are helping to transform the way people live and work.

Our core technology, built from years of research on IoT sensing with smart cameras and RF devices, has been recognized with multiple awards for its outstanding impact on the industry.

Our focus is on improving operational efficiency and enhancing the customer experience in the cities we serve. Our goal is to help transform Hong Kong into a world-class smart city, starting with our recent target market in smart property management across various sectors such as shopping malls, exhibition halls, hotels, museums, and hospitals.

We are incubated by HKSTP and continuously strive to expand our expertise in RF sensing, while aiming to further extend our sensing capabilities through AI-powered smart cameras. At Smart Sensing Ltd., we believe in the power of technology to make our cities smarter, safer, and more connected. Join us on our mission to shape the future of data gathering and analysis.

Contact Us：info@smartsensing.biz

Our DaaS-platform for Industrial AIoT Solutions

Core Components

Local Sensors & Actuators

Well-optimized infrastructure for seamless communications between sensors and actuators for sensing (i.e. Wi-Fi, Bluetooth, and visual data etc.)

Computing Cloud

Exclusive algorithms for recovering trajectory under virtual MAC, IR/Thermal camera video analytics, Machine learning and federated learning techniques for multi-modal data, Multi-variate data forecasting…

Intelligent Insight/Control

Super user-friendly interfaces for target users to manipulate over the systems

Edge AI

Our assets and sensors at the Edge can be trained to perform autonomous tasks. Real-time data analytics and remote management of devices can be done in the field to achieve optimised performance and cost efficiency.

Application Modules

Smart Indoor Sensing Solution

iBeacon-based

Passive Positioning

Installing iBeacon seeds on site, Mobile apps then receive the iBeacon signals and interpret the changes of RF intensity to determine its distance, location and movement speed.

Results & Functions:

01. Precise Location
– Enable indoor navigation
– Apps / action trigger when entering certain zone

02. Relative Speed
– Enable real-time indoor speed detection

Access-Point-based

Active Positioning

Installing a modified firmware into local Access-Point device (i.e. Router), to detect the mac address and RF intensity of mobile devices, then determine the relative position of those mobile devices.

Results & Functions:

01. Crowd Distribution
– Enable heat map visualization for crowd location

02. Mobility Tracking
– Tracing the pattern of mobile device

The Smart Edge AI Crowd Counting System

The Smart Edge AI Crowd Counting System combines edge computing and artificial intelligence to enable processing at the device level, using machine learning to offer crowd-counting and line-crossing counting features. This smart-city solution is designed for real-time monitoring and analysis of pedestrian traffic for large scale events, such as New Year’s fireworks display, in a portable and flexible approach of hardware. It can be adapted from a mobile phone cam, a high-resolution cameras and even aerial drone devices for enhanced functionality. The result of all edge devices sends to the IoT platform for transforming the data of multiple investigation points to the exact crowd information of the scene(s).

This system leverages deep learning algorithms and computer vision technologies, such as Observation-Centric SORT (OC-SORT) and ByteTrack, to enhance the robustness of multi-object tracking by addressing the limitations of traditional Kalman filter-based approaches and also accurately count/track the number of people in specific areas. The system introduces two key innovations:

1) It reduces accumulated errors when objects are untracked via incorporating virtual observations from historical time steps to prevent error accumulation. These virtual observations are generated using the last-seen observation before the object was untracked and the latest observation that reactivates the track.

2) It considers the direction consistency of object motion. By incorporating direction consistency, OC-SORT improves the accuracy of object association, even in the presence of noise in direction estimation. It achieves high frame rates on a single CPU.

The Smart Edge AI Crowd Counting System is a fast and easy-to-set-up solution, where the minimal requires a smart phone with cam, a GPU laptop & cam stands. By leveraging different hardware devices in conjunction with optimized algorithms, it provides accurate and efficient people counting across various scenarios. This system/solution provides an easier and portable way to enhance visitor safety via crowd density/line crossing detection and provide several benefits –

Improved safety: they can help monitor the number of people in a particular area, which can be useful for ensuring that public spaces are not overcrowded. This can help prevent accidents and emergencies, particularly in areas where large crowds gather.
Enhanced security: they can provide real-time monitoring of public spaces. This can help identify potential security threats and provide a quick response in case of an emergency.
Better resource allocation: they can help organisers and security guards better understand how people use public spaces. This information can be used to allocate resources more efficiently, such as by adjusting staffing levels or improving infrastructure.
Improved planning: they can be used to inform crowd control planning decisions, such as where to allocate more manpower or how to improve controlling the flow of crowd. This can help ensure that the community’s needs are being met and that resources are being used wisely.